摘要
VAR是一种测定和控制金融风险的量化模型,本文讨论非参数统计方法在其中的应用。从历史模拟法这种非参数VAR模型的基本思想出发,提出一种以对金融资产收益率分布的核密度估计为基础的VAR模型,并将其与RiskMetrics这种参数VAR模型做了对比,结果表明本文的方法有效避免了RiskMetrics所存在的问题。
VAR(value-at-risk)is a quantitative model designed for the measurement and control of financial risk. This paper discusses the application of non-parametric statistics in this model. We follow the idea of Historical Simulation-a typical nonparametric VAR model and construct a new model based upon the kernal density estimation of the distribution of asset's returns. We compare the new model with RiskMetrics-a typical parametric model and the results show that the new model effectively avoids some un-desired effects of RiskMetrics.
出处
《系统工程》
CSCD
1999年第5期25-32,共8页
Systems Engineering
